Neurofibrillary tangles are one of the primary hallmarks of Alzheimer disease, and closely correlate with clinical severity. They consist of hyperphosphorylated tau that aggregates in the neuronal cell body. Cross sectional histological studies suggest that tangles kill neurons. Tau disaggregation is thus a major target of the pharmaceutical industry, with at least one drug already in human clinical trials. However, recent data suggest that tangles may not be toxic. We propose to test an alternative hypothesis, in which soluble tau induced toxicity leads to activation of apoptosis related cascades, followed by tangle formation. In order to test this model, we have developed in vivo multiphoton based longitudinal imaging techniques for tangles, caspase activation, and propidium iodide (a marker of loss of membrane integrity). Aim 1 applies this approach to Tg4510 P301L tau mice that develop NFT and neuronal loss. Surprisingly, we see caspase activation that appears to precede and predict tangle formation. We will also use array tomography, an advanced microscopic method using ultrathin sections of tissue, to determine the characteristics of the caspase positive, tangle negative cells to test the hypothesis that kinase activation leads to phosphorylation of tau, and that caspase activation leads to the truncation of tau [unreadable]the combination of which causes tau aggregation into a tangle. We will use gene transfer and pharmacological studies to further dissect this pathway, to ask whether caspase activation, and caspase truncation of tau, are necessary or sufficient to cause tangle formation, neuronal distress or death. Aim 2 asks if NFT remain in viable cells and are long lived, or if they are toxic to the neurons in which they are found either in terms of structure or function. These experiments take advantage of the power of longitudinal in vivo multiphoton imaging to follow the fate of individual tangle bearing neurons for weeks to months. We will also test the hypothesis that tangle bearing neurons are excluded from participation in normal neural system activation from physiological stimuli by exposing animals to an enriched environment, then examining individual neurons in hippocampal subfields for immediate early gene (Arc) expression and the presence or absence of tangles. Aim 3 will rigorously test the hypothesis that wild type, nonmutant tau, undergoes similar phenomenon and that the toxic effects of tau over expression are due, in large part, to soluble tau. We take advantage of 3 models: examination of an alternative transgenic model, the hTau mice which express a minigene of wild type human tau (on a tau null background), introduction of wild type tau (or truncated forms of tau) into wild type mice using gene transfer approaches with AAV2 gene vectors, and the Tg4510 mice, which harbor a tet-response element driving the tau gene. These models allow comparison of wild type and mutant tau, and will also allow us to distinguish te effects of soluble tau from those of misfolded or aggregated tau. Taken together, our proposed studies will test the hypotheses that a non-tangle related mechanism of tau toxicity initiates apoptotic cascades, and that tangles may be a relatively nontoxic, long lived species. The results will have direct impact on design of therapeutic agents destined for clinical trials.